Spiking Neural Networks for Robotic Applications

Abstract

The biologically inspired spiking neural network can be said to mimic the most neural network models in existence that is evolved from artificial neural networks. This concept was derived from the nervous system and was able to generate electric impulses, commonly known as spikes or action impulses. Here the neural models try to replicate the biological neurons almost accurately and can be considered to be more powerful than its peers as it was able to integrate temporal information. As such they can said to have great potential in several complex applications like classification, mapping, and pattern recognition, etc. Out of the available spiking neuron models, Leaky-Integrate-and-Fire is very frequently applied. Spiking neural networks are gaining rapid importance in the last few years following the sharp incline in artificial intelligence field. In this modern era robots are incorporated in our daily life, and said to have the potential to increase economic growth and productivity.